Heparin-mimicking polyethersulfone membranes – hemocompatibility, cytocompatibility, antifouling and antibacterial properties

Author’s Accepted Manuscript Heparin-mimicking polyethersulfone Membraneshemocompatibility, cytocompatibility, antifouling and antibacterial properties Shuang-Si Li, Yi Xie, Tao Xiang, Lang Ma, Chao He, Shu-dong Sun, Chang-Sheng Zhao www.elsevier.com

PII: DOI: Reference:

S0376-7388(15)30216-7 http://dx.doi.org/10.1016/j.memsci.2015.09.054 MEMSCI14010

To appear in: Journal of Membrane Science Received date: 3 August 2015 Revised date: 18 September 2015 Accepted date: 20 September 2015 Cite this article as: Shuang-Si Li, Yi Xie, Tao Xiang, Lang Ma, Chao He, Shudong Sun and Chang-Sheng Zhao, Heparin-mimicking polyethersulfone Membranes-hemocompatibility, cytocompatibility, antifouling and antibacterial p r o p e r t i e s , Journal of Membrane Science, http://dx.doi.org/10.1016/j.memsci.2015.09.054 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

Heparin-mimicking Polyethersulfone Membranes - Hemocompatibility,

2

Cytocompatibility, Antifouling and Antibacterial properties Shuang-Si Lia, Yi Xiea, Tao Xianga, Lang Maa, Chao Hea, Shu-dong Suna and

3

Chang-Sheng Zhaoa,b*

4 5

a

6

Materials Engineering, Sichuan University, Chengdu 610065, People’s Republic of

7

China

8

b

9

Chengdu 610064, China

10

*Corresponding author.

11

E-mail address: [email protected] or [email protected]

12

Tel.: +86-28-85400453; Fax: +86-28-85405402.

13

Abstract

College of Polymer Science and Engineering, State Key Laboratory of Polymer

National Engineering Research Center for Biomaterials, Sichuan University,

14

In this study, a series of heparin-mimicking polyethersulfone (PES) membranes

15

were prepared through a highly efficient, convenient and universal in situ

16

cross-linking polymerization technique coupled with a phase inversion technique.

17

Two kinds of monomers, sodium acrylate (AANa) and sodium styrene sulfonate

18

(SSNa) were used to introduce functional carboxyl and sulfonic groups onto PES

19

membrane surfaces, respectively; and thus to mimic the chemical structure and

20

biological activity of heparin. The heparin-mimicking membranes showed decreased

21

protein adsorption, greatly suppressed platelet adhesion (decreased by more than

22

93%), and prolonged clotting times (prolonged as much as 60s for APTTs and 20s for

23

TTs, respectively) compared to pristine PES membrane, which confirmed the 1

1

enhanced blood compatibility of the modified membranes. The cell culture and 3-(4,

2

5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assays revealed that

3

the heparin-mimicking membranes had a favorable trend in terms of endothelial cell

4

proliferation and cell morphology. Moreover, the membranes showed good

5

antifouling property. These results confirmed that the highly efficient and convenient

6

in-situ polymerization had endowed the heparin-mimicking membranes with excellent

7

biocompatibility, which might have great potential application in blood purification

8

fields. In addition, the membranes were loaded with Ag nanoparticles, for which

9

exhibited significant inhibition capability for Escherichia coli and Staphylococcus

10

aureus, and thus confirmed the versatility of the protocol.

11

Keywords: In situ cross-linking polymerization; Polyethersulfone membrane;

12

Heparin-mimicking; Biocompatibility; Antibacterial property

13

Symbol

Meaning (common units)

Jv

water flux of membranes (mL/h·m2·mmHg)

V

volume of the permeated solution (mL)

T

time

S

effective membrane area (m2)

ΔP

pressure applied to the membrane (mmHg)

JRR

flux recovery ratio of the membranes

Jv1

the PBS flux before protein ultrafiltration experiment (mL/m2·h·mmHg)

Jv2

the PBS flux after protein ultrafiltration experiment (mL/m2·h· mmHg) 2

Cp

BSA concentration in the permeated solution (mg/mL)

Cb

BSA concentration in the bulk solution (mg/mL)

R

protein rejection ratio

1 2

1. Introduction

3

Heparin, a linear polysaccharides consisting of repeating disaccharide units of 1,

4

4-linked uronic acid (D-glucuronic(GlcA) or L-iduronic acid(IdoA)) and D

5

-glucosamine (GlcN), has a higher negative charge density than any other known

6

biological macromolecules due to the presence of negatively charged carboxyl and

7

sulfonic group[1, 2]. It is capable of interacting with coagulation factors XIa, IXa, Xa,

8

and IIa (thrombin), and has been widely used as anticoagulant reagent. However, it is

9

difficult to directly use heparin as anticoagulant material to enhance the

10

hemocompatibility of polymeric membranes due to its water solubility; but many

11

studies have been carried out on surface heparinization, such as blending[3], surface

12

grafting and surface coating[4, 5], as well as layer-by-layer assembly [6]. These

13

heparin-modified membranes showed improved hemocompatibility due to the

14

bioactivity of heparin. However, as a product derived from animals, direct utilization

15

of heparin for membrane modification does exist some drawbacks, e.g. the high cost

16

of heparin inhibits its large-scale use for membrane surface modification; moreover, a

17

dramatic loss of bioactivity and degradation in vivo will occur in biological systems

18

due to the covalent or noncovalent bindings with blood components etc. which lead to

19

the lack of stability and durability [3, 7]. All these may prevent its practical 3

1

application in biomedical devices. Therefore, it is of great importance to find an

2

alternative to be used for modifying biomedical membranes.

3

It is considered that the anticoagulant activity of heparin is mainly caused by the

4

existence of the carboxyl and sulfonic groups on the backbone [2]. Accordingly, great

5

efforts had been made to design and synthesize heparin-mimicking polymers,

6

containing sulfate, sulfamide and carboxylate groups [2, 8]. The synthetic

7

heparin-mimicking polymers showed some outstanding advantages, such as

8

anticoagulant ability [9], mediated inflammation [10], and promotion of cell adhesion

9

and proliferation by binding and stabilization of cell growth factors [8, 11]. Due to the

10

excellent advantages, the heparin-mimicking polymers were used to design and

11

prepare polymeric membranes to improve the biocompatibility [12-15]. After

12

introducing the functional groups, the membranes showed enhanced biocompatibility.

13

In our recent studies, several heparin-mimicking polymers were synthesized for the

14

modification of polyethersulfone (PES) membranes. However, due to the poor

15

miscibility between the polymers and PES, the blended amounts of the polymers into

16

PES were limited. In order to solve this problem, heparin-mimicking PES was

17

designed for improving the blood compatibility of PES membranes [16-18]. However,

18

the above methods were sometimes limited, since the synthesis of the polymers was a

19

time-consuming complicated process [19-21].

20

To further develop the physical blending method and allow it to be more suitable

21

for industrial applications, we recently provided a method termed “in situ

22

cross-linking

polymerization/copolymerization” 4

[22-24].

Due

to

the

1

semi-interpenetrating network generated during the polymerization/copolymerization,

2

the obtained blending system showed excellent miscibility; and the resulted

3

membranes showed no clear phase separation and displayed not only better blood

4

compatibility but also a good mechanical property [25, 26].

5

In this study, inspired by the above heparin-mimicking concept and in situ

6

cross-linking polymerization method, a simple and convenient method to introduce

7

carboxyl and sulfonic groups into PES membranes by in situ cross-linking

8

polymerization was carried out and the membrane performances were explored. PES

9

was selected as a membrane matrix because of its good oxidative, thermal, and

10

hydrolytic stabilities, as well as good mechanical and film-forming properties; and

11

had been widely applied in the fields of artificial organs and medical devices [27-31].

12

Sodium acrylate (AANa) and sodium p-styrene sulfonate (SSNa) were selected as the

13

functional monomers because of their high reaction activity in free radical

14

polymerization. Then, a series of heparin-mimicking membranes were prepared by a

15

phase inversion technique. Furthermore, the Ag nanoparticles (Ag NPs) were

16

embedded in the membranes to endow with antibacterial property, when considered

17

the membranes for long-time using in future portable hemodialyzer (as shown in

18

Scheme.S1). The chemical components, surface and cross-section structure of the

19

membranes were confirmed by ATR-FTIR, element analysis and scanning electron

20

microscopy (SEM). While the hemocompatibility was explored by wettability, static

21

protein adsorption, platelet adhesion, activated partial thromboplastin time (APTT),

22

thrombin time (TT). Furthermore, human vessel endothelial cells (HUVECs) were 5

1

used as model cells to investigate the cell viability of the membranes. The

2

antibacterial property was tested via bacterial inhibition zone towards E. coli and S.

3

aureus, respectively [32, 33].

4 5

Scheme.1 Preparation process and multi-functionality of the heparin-mimicking

6

membranes.

7

2. Experimental Section

8

2.1. Materials

9

Poly (ether sulfone) (PES, Ultrason E 6020P) was purchased from BASF chemical

10

company (Germany). Sodium 4-vinylbenzenssulfonate (SSNa，90%) and sodium

11

acrylate (AANa) were purchased from Aladdin Reagent Co. Ltd. (China). N,

12

N'-methylene bisacrylamide (MBA，98%) and azoisobutyronitrile (AIBN，99%) were

13

obtained from the Chemical Reagent Factory of Kelong, China. The solvent N,

14

N-dimethylacetamide (DMAc) was distilled under reduced pressure with calcium

15

hydride (CaH2) to remove the water. Bovine serum albumin (BSA) and bovine serum

16

fibrinogen (FBG) were obtained from Sigma Chemical Co. Micro BCATM protein 6

1

assay reagent kits were the products of PIERCE. APTT and TT reagent kits were

2

purchased from SIEMENS. All the other chemicals (analytical grade) were obtained

3

from the Chemical Reagent Factory of Kelong, China, and used without further

4

purification. More detailed information for other materials was included in the

5

Supplementary Materials.

6

2.2. Preparation of heparin-mimicking membranes

7

In this paper, two kinds of monomers, AANa and SSNa, were used to introduce

8

carboxyl and sulfonic groups, respectively. To prepare casting solution, PES and the

9

monomer (NaAA or SSNa) were separately dissolved in a 250 mL three-necked round

10

flask with appropriate amount of DMAc until a homogeneous solution was obtained.

11

After pumping and aerating with nitrogen for three times, a mixture of AIBN and

12

MBA was added into the flask (detailed components are presented in Table.1). Then

13

the polymerization was carried out with mechanically stirring (500 rpm) under

14

nitrogen at 75 °C for 24 h, the polymerization was then exposed to air to terminate the

15

reaction. The obtained two kinds of solutions were named APES (for AANa and PES)

16

and SPES (for SSNa and PES), respectively. Finally, different kinds of casting

17

solutions were prepared as shown in Table 1.

18

All the membranes were prepared as the following procedures. After 20-minute

19

degassing, the casting solution was spin-coated on a glass surface, which was then

20

immersed into deionized water and kept for 5 minutes, and a thin membrane with a

21

controlled thickness of (70±5) μm was prepared. Then, the membrane was kept in

22

deionized water for 48h to remove the residual DMAc. 7

1

The white opaque membranes obtained via the phase inversion method were named

2

M-0-0, M-2-0, M-2-1, M-2-3, M-0-2, M-1-2 and M-3-2 (the compositions are shown

3

in Table.1), respectively. The membrane M-2-2 was not studied in this paper, since we

4

just want to find out the relationship between the heparin-mimicking content and the

5

membrane performance, which could be used to determine the optimum formula.

6

Table.1. Weight compositions of the cast solutions for the prepared membranes Sample M-0-0 M-2-0 M-2-1 M-2-3 M-0-2 M-1-2 M-3-2

7

PES/wt. % 16 14 13 11 14 13 11

APES/wt. % 0 2 2 2 0 1 3

SPES/wt. % DMAc/wt. % 0 84 0 84 1 84 3 84 2 84 2 84 2 84

2.3. Characterization of the prepared heparin-like membranes

8

The surface compositions of the membranes were investigated by attenuated total

9

reflection-Fourier transform infrared (ATR-FTIR) spectra on a Nicolet-560

10

spectrophotometer (Nicol, US) between 3000 cm-1 and 750 cm-1, with a resolution of

11

2cm-1. It is difficult to confirm the successful introduction of the sulfonic groups into

12

PES membrane since there are also S elements in PES backbone. Thus FTIR spectra

13

of the resulted polymer solution (SPES) were measured and the results were

14

compared with poly (sodium 4-styrenesulfonate) (Mw~70000) (the detail information

15

was presented in Supplementary Materials). NMR spectra were not used to prove the

16

presence of the sulfonic groups on PES substrate (the detail explanation was present

17

in Supplementary Materials, page 2.).

18

To obtain the cross-section and surface morphologies for the membranes, a 8

1

scanning electron microscopy (SEM) (JSM- 7500F, JEOL, Japan) was used. After

2

freeze-drying overnight, the membrane samples were quenched and fractured in liquid

3

nitrogen, attached to the sample supports and coated with gold layers.

4

A contact angle goniometer (OCA20, Data physics, Germany) equipped with a

5

video capture was applied to characterize the hydrophilicity/hydrophobicity of the

6

membranes. 3μL of water was dropped on the surface of the membrane at room

7

temperature with an automatic piston syringe and photographed. At least three

8

measurements were averaged to get a reliable value.

9

2.4 Blood compatibility

10

2.4.1 Plasma collection

11

Healthy human fresh blood (man, 25 years old) was collected using vacuum tubes

12

(5mL, Jiangsu Kangjian Inc., China) containing sodium citrate as anticoagulant

13

(anticoagulant-to-blood ratio, 1:9 v/v). The blood was centrifuged at 1000 rpm for 15

14

min to obtain platelet-rich plasma (PRP) or at 4000 rpm for 15 min to obtain

15

platelet-poor plasma (PPP). The same donor blood samples were used all through the

16

blood tests.

17

In this paper, in order to explore the blood compatibility in vitro, static protein

18

adsorption, platelet adhesion, activated partial thromboplastin time (APTT) and

19

thrombin time (TT) were conducted, since these indexes were widely used to evaluate

20

the in vitro blood compatibility.

21

2.4.2. Static protein adsorption

22

Protein adsorption, which is considered as the key step when a material comes in 9

1

contact with blood [34]. In this study, protein adsorption experiments were conducted

2

with BSA and FBG solutions under static condition. Firstly, the membrane with an

3

area of 1 cm ×1 cm was pretreated with phosphatic buffer solution (PBS, pH＝7.4) at

4

4 °C for 24 h. Then the membrane was immersed in PBS containing BSA or FBG

5

with a concentration of 1 mg/mL, and incubated at 37 °C for 1 h. After protein

6

adsorption, the membrane was slightly rinsed with PBS and deionized water

7

sequentially. Then, the membrane was immersed in a solution (2 wt. % of sodium

8

dodecyl sulfate (SDS) aqueous solution) at 37 °C, and shaken for 2h to remove the

9

adsorbed protein [35]. The protein in the SDS solution was determined using the

10

Micro BCA™ Protein Assay Reagent Kit (PIERCE), and the adsorbed protein amount

11

was calculated. At least three testing results were averaged to obtain a reliable value.

12

2.4.3. Platelet adhesion and activation

13

The experiment procedure was the same as described in our previous works [16,

14

35]. The membrane with an area of 1 cm×1 cm was immersed in PBS and equilibrated

15

at 37 °C for 1 h;

16

was removed with a dropper, and the membrane was mildly washed three times in

17

PBS, then the platelets adhered on the membrane were fixed with 2.5wt%

18

glutaraldehyde in PBS at 4°C for 24h. Finally, the sample was rinsed in PBS, and then

19

subjected to a drying process by immersing in a series of graded alcohol-PBS (30, 50,

20

70, 80, 90, 95 and 100 %) for 15 min. The platelets adhered on the membrane surface

21

were observed using an SEM (JSM-7500F, JEOL, Japan). Then, the number of the

22

adherent platelets on the membrane surface was calculated from five SEM pictures at

and then incubated in 1 mL fresh PRP at 37 °C. After 1 h, the PRP

10

1

500× magnification from different places on the same sample. Furthermore, the

2

platelet activation was characterized by commercial enzyme-linked immunosorbent

3

assays (ELISA) (Human Platelet Factor 4 (PF4), Cusabio Biotech Co. Ltd., China)

4

and the detailed information was presented in Supplementary Materials.

5

2.4.5. Activated partial thromboplastin time (APTT) & thrombin time (TT) test

6

For testing the antithrombogenicity of the membranes, activated partial

7

thromboplastin time (APTT) and thrombin time (TT) were measured by a

8

semi-automatic blood coagulation analyzer CA-50 (Sysmex Corporation, Kobe,

9

Japan).

10

The testing process referred to our previous works [36, 37] was as follows: the

11

membranes (0.5 cm×0.5 cm for each) were immersed in a 96-well plates with 200μL

12

PBS in each well at 4 °C for 24 h, then the PBS was removed and 100μL PPP was

13

introduced. After incubating at 37 °C for 30 min, 50μL of the incubated PPP was

14

added into a test cup, followed by the addition of 50μL of APTT agent (incubated at

15

37 °C 10min before using). After incubating at 37 °C for 3 min, 50μL of 25 mM

16

CaCl2 solution was added, and then the APTT was measured. At least three

17

measurements were averaged to get a reliable value, and the results were analyzed by

18

statistical method. For the TT test, 100μL of TT agent was added into the test cup

19

(containing 50μL of the incubated PPP), and then the TT was measured.

20

2.5. Cytocompatibility

21

2.5.1. Cell culture

22

To explore the cytocompatibility of the membranes, human umbilical vein 11

1

endothelial cells (HUVECs) were used as model cells to culture in R1640 medium

2

supplemented with 10 % fetal bovine serum (FBS, Hyclone, USA), 2mM L-glutamine,

3

and 1 % (V/V) antibiotics mixture (10,000 U penicillin and 10mg streptomycin), since

4

the endothelial layer was important for maintaining safety of blood-contact devices

5

and was involved in prevention of excessive tissue ingrowth (intimal hyperplasia) and

6

thrombogenesis. The culture was maintained in a humidified atmosphere of 5 % CO2

7

at 37 °C (Queue Incubator, Paris, France), and the culture medium was changed every

8

day. Confluent cells were detached from the culture flask with sterilized PBS and

9

0.05 % trypsin/EDTA solution.

10

The PES and modified PES membranes were cut into pieces (1 cm×1 cm for each)

11

and placed in a 24-well cell culture polystyrene plate and then sterilized by

12

γ-irradiation. Then, HUVECs with a density of ~2.5×104 cells/cm2 were seeded and

13

cultured in the incubator for predetermined time.

14

2.5.2. Cell morphology observation

15

After 6 days incubation, the membranes were rinsed gently with sterilized PBS;

16

then the membranes were stained with a FDA/PI mixed solution for fluorescence

17

microscope

18

6-diamidino-2-phenylindole (DAPI) for laser scanning confocal fluorescence

19

microscope (LSCF, Leica, Switzerland) according to the instruction manuals,

20

respectively.

21

2.5.3. 3-(4, 5-dimethylthiazol-2-yl)-2, 5-diphenyltetrazolium bromide (MTT) assay

22

observation

and

rhodamine-conjugated

phallodin

and

4a,

After cell culture for 2, 4 and 6 days, the viability of the hepatocytes was 12

1

determined by MTT assay. In a typical MTT assay test, the dehydrogenase in the

2

mitochondria of living cells was able to convert MTT reagent into a blue formazan

3

crystalline product, which can be dissolved in water or ethanol and then monitored

4

using a Microplate reader (model 550, Bio-Rad) at 492 nm. All the experiments were

5

repeated three times, and the results were expressed as mean ± SD. The statistical

6

significance was assessed by Student’s t-test, with the level of significance set at p <

7

0.05.

8

2.6. Protein antifouling property

9

Protein antifouling property, as a key aspect of blood-contacting material, should

10

draw our attention when designing hemodialysis membrane [35, 36, 38]. Thus,

11

ultrafiltration of BSA solution through the membrane was carried out to investigate its

12

antifouling property. Firstly, BSA solution was prepared by dissolving BSA in PBS

13

(pH=7.4) with a concentration of 1.0 mg/mL. Then the membrane was placed in a

14

dead-end ultrafiltration cell with an effective membrane area of 3.9 cm2. For the test,

15

the membrane was firstly pre-compacted by PBS (pH = 7.4) at a pressure of 0.07 MPa

16

for 20 min to reach a steady flux. Then the pressure was adjusted to 0.06 MPa, and the

17

PBS flux within 5 minutes was measured (and this measurement was repeated for 5

18

times). After the filtration of PBS, the feed solution was switched to 1.0 mg/mL BSA

19

solution, and the operation was the same as that for the PBS ultrafiltration. After the

20

BSA solution filtration, the membrane was immersed in PBS for 1 h, and then the

21

above process was repeated again. The fluxes (Jv) of PBS and BSA solution through

22

the membrane were measured. (The detailed information was presented in 13

1 2

Supplementary Materials). The protein rejection ratio (R) was defined as the following: Cp

3

R  (1 

4

where Cp and Cb represent the protein concentrations of the permeated and bulk

5

solutions, respectively. Then, five R values during the time of 30-50min were

6

averaged to obtain a mean value. The protein concentration was measured by an UV–

7

vis spectrophotometer at the wavelength of 278 nm. After protein filtration, the

8

membrane was cleaned with deionized water. Then, the flux of the cleaned membrane

9

was measured again and the flux recovery ratio (JRR) was calculated by the following

Cb

) 100%

(1)

10

equation:

11

J RR 

12

where Jv1 and Jv2 are the PBS fluxes before and after each protein ultrafiltration,

13

respectively.

14

2.7. Antibacterial experiment

Jv2 100% J v1

(2)

15

For portable hemodialyzer used in future, the hollow fiber membranes with better

16

anti-coagulant inner layer and anti-bacterial outer layer were required (as shown in

17

Scheme.S1), as they need to effectively work for a long time at a low cost. For this

18

purpose, Ag nano-particles were introduced into the membranes and the antibacterial

19

activity was tested. According to our previous work [39], the negatively charged

20

carboxyl groups in the membranes could be used for loading Ag nanoparticles, and

21

then endowed with antibacterial activity. Firstly, the membranes with 2 wt. %, 4 wt. % 14

1

and 6 wt. % AANa were prepared and named as M-2-0, M-4-0 and M-6-0 in the same

2

way as described above. After immersing in 20 mL 0.1 M AgNO3 solution and stirring

3

at 25 °C in the dark for 12 h, the membranes were took out and washed with DI water

4

to remove the excess AgNO3. Then the membranes were immersed in 0.1 M NaBH4

5

solution and stirred at 25 °C in the dark for another 12 h; and then the membranes

6

were stored in DI water for 2 days. The characterization of the Ag NP embedded

7

membranes were shown in the Supplementary Materials (Fig.S4 and Fig.S5).

8

Escherichia coli (E. coli, gram negative) and Staphylococcus aureus (S. aureus,

9

gram positive) bacteria were used as the model bacteria to evaluate the antibacterial

10

characteristics and bactericidal efficacy for the membranes. And the detailed

11

operations were referred to Xia’s work [39].

12

3. Results and discussions

13

3.1. Characterization of the membranes

14

ATR-FTIR was used to characterize the membranes; the spectra for the membrane

15

samples are shown in Fig. 1. As shown in the figure, after introducing poly (AANa) in

16

the membranes, a new peak at 1665 cm-1 attributed to the –C=O stretching was

17

observed, indicating the successful introduction of AANa and the presence of the

18

functional groups on the membrane surfaces. The absorption peaks observed at 1072

19

and 1011 cm-1 were ascribed to the symmetric stretching vibrations and the

20

asymmetrical stretching vibrations of the highly polar –SO3− groups of poly (SSNa),

21

respectively [40]. With increasing the amount of the poly (SSNa), the peak intensity

22

increased. For further confirming the successful introduction of poly (SSNa) onto the 15

1

PES membrane surface, supplementary FTIR spectra of the resulted SPES solution

2

and purchased poly (sodium 4-styrenesufonate) (with Mw~70000) were presented in

3

Fig.S1 in Supplementary Materials.

4 5

Fig.1 ATR-FTIR spectra for the membranes

6

The cross-sectional SEM micrographs of the membranes were shown in Fig. 2 (a).

7

As observed in the figure, a dense skin layer and a porous sub-layer with a finger-like

8

structure were observed. It was also found that the skin layers of the

9

heparin-mimicking membranes were thinner than that of the pristine one. What's more,

10

the pore diameter and porosity of the finger-like structure became larger with

11

increasing the fraction of the hydrophilic components. The results were similar to

12

those of the membranes prepared by Han et al [22]. The SEM images suggested that

13

the structure of the membrane had been altered owing to the modification. With the

14

addition of a high molecular weight polymer into a casting dope, the pore size in the

15

sub-layers became bigger and the morphology would change as the PES membrane

16

became more hydrophilic [41]. It could be observed from Fig. 2(b) that the surfaces of

17

the heparin-mimicking membranes became rougher than the pristine PES membrane, 16

1

which might be caused by the phase separation during membrane formation [18].

2

When the hydrophilic component reached 5 % (as for M-2-3 and M-3-2), some

3

micropores were observed, which would affect the membrane protein antifouling

4

property as discussed in the following section.

5 6

Fig.2 SEM micrographs for (a) the cross-section views and (b) the surface views of

7

the membranes. (Magnification: ×1000; Scale bar: 10μm)

8

Water

contact

angle

(WCA)

was

usually

used

to

detect

the

9

hydrophilicity/hydrophobicity of a material surface, which then provided information

10

about the wettability property of the material surface and the interaction energy

11

between the material surface and the liquid [42, 43]. Many factors such as the

12

hydrophilicity (or hydrophobicity), roughness, porosity, pore size, and its distribution 17

1

would affect the water contact angle data [44].

2

The WCA results of the membranes were shown in Fig.3. As shown in the figure,

3

the WCAs of the heparin-mimicking membranes decreased to some extent compared

4

to the pristine PES membrane; and further decreased with increasing the hydrophilic

5

components. These proved that the surface hydrophilicity had been improved after the

6

modification, which also confirmed the successful introduction of the carboxyl groups

7

and sulfonic groups. The capillarity of phase-inversion membranes might also have

8

effect on decreasing WCA. To make the illustration clear, further study on the

9

enhanced surface hydrophilicity using evaporation membranes could find in the

10

Supplementary Materials (Fig.S2).

11 12

Fig.3 Static water contact angle for the pristine and heparin-mimicking PES

13

membranes

14

3.2. Blood compatibility

15

3.2.1. Static protein adsorption

16

The amount of protein adsorbed onto a material surface was reported as one of the

17

most important factors when evaluating the hemocompatibility of biomaterials. In this 18

1

study, both the BSA and BFG adsorptions were tested, and the results are presented in

2

Fig.4. As shown in the figure, the protein adsorption amounts to the

3

heparin-mimicking PES membranes decreased slightly compared to the pristine PES

4

membrane. As reported previously [38], the decreased protein adsorption was resulted

5

from the hydration layer formed between water and the heparin-mimicking groups

6

(-COO- and -SO3-) on the surface. In addition, both BSA and BFG contain net

7

negative charges in PBS (pH=7), thus the electrostatic repulsion between the protein

8

and heparin-mimicking membranes would also contribute to the decreased protein

9

adsorption amount. The improved anti-protein adsorption property might enhance the

10

hemocompatibility of the heparin-mimicking membranes, which would be discussed

11

in the following sections.

12 13

Fig.4 BSA and FBG adsorption for the pristine and heparin-mimicking PES

14

membranes

15

3.2.2. Platelet adhesion

16

According to the report of Grunkemeier and coworkers [45], platelet morphology

17

change and activation after adhering onto material surfaces can be classified into five 19

1

stages: discoid, dendritic (early pseudopodia), spread/dendritic (intermediate

2

pseudopodia), spreading (late pseudopodia and hyaloplasm spreading), and fully

3

spreading (hyaloplasm well spread but no distinct pseudopodia) [45, 46]. When an

4

incompatible material comes in contact with blood, the platelets are initiated and

5

inclined to fully spread to achieve the largest area coverage on the material surface

6

[26]. Thus, the number and morphology of the adhering platelets for the membranes

7

were studied. In addition, platelet activation (PF4) was evaluated via commercial

8

enzyme-linked immunosorbent assays (ELISA) and the results were shown in Fig.S3

9

in Supplementary Materials.

10

Fig.5 represented the SEM images of the platelet adhesion for the membranes. By

11

comparing these figures, it was observed that numerous platelets were aggregated and

12

accumulated on the pristine PES membrane surface; and the platelets spread in

13

flattened and irregular shapes, and a lot of pseudopodia were observed, which

14

indicated that platelet activation might occur on the surface of the PES membrane and

15

this was an undesirable phenomenon for clinical hemodialysis. However, for the

16

heparin-mimicking membranes, the adhered platelets were much less observed and

17

the platelets expressed rounded morphology with nearly no pseudopodium and

18

deformation. From these results, it was concluded that the modification of the

19

membranes could reduce the platelet adhesion and inhibit the platelet activation.

20

1 2

Fig.5 (a) SEM images of the platelets adhering onto the membranes, images (a-g)

3

were the partial enlarged ones for images (A-G); (b) the number of the adhering

4

platelets onto the membranes (from platelet-rich plasma estimated by SEM images)

5

3.2.4. Activated partial thromboplastin time (APTT) & thrombin time (TT)

6

There exist three pathways in coagulant system when materials come in contact

7

with blood: the intrinsic pathways, the extrinsic pathway, and the common pathway

8

[16]. Heparin-mimicking materials can prolong the blood clotting time [17] due to

9

carboxyl and sulfonic groups, which are available for binding of coagulation factors

10

[14]. APTT is always used to measure the inhibited efficacy of both the intrinsic

11

(sometimes referred to as the contact activation pathway), and the common

12

coagulation pathway. While, the time taken for the thrombin fibrinogen converted into

13

fibrin in the PPP is often evaluated by TT test. The faster conversion of fibrinogen

14

(which is indicated by a shorter clotting time) indicates that thrombus will easily

15

formed [47]. 21

1

Fig.6 showed the results of APTT and TT for the membranes. It was observed that

2

the APTTs and TTs for the modified membranes prolonged compared to pristine PES

3

membrane. In addition, with the increase of the heparin-like functional group contents,

4

the APTTs and TTs increased. For the M-2-3 and M-3-2 membranes, the APTTs

5

increased nearly 150 % and 130 %, respectively; and the TTs for the two samples

6

increased approximately 95 % and 62 %, respectively, compared to the pristine PES

7

membrane. In a word, the facile heparin-mimicking modification had endowed the

8

membranes with excellent blood compatibility.

9 10

Fig.6 APTTs and TTs for the pristine PES and the heparin-mimicking membranes

11

(Values are expressed as mean ± SD. P#, P* < 0.05 compared with plasma (PPP) and

12

pristine PES membrane, respectively.)

13

3.3．Cytocompatibility

14

An endothelial layer was important for maintaining safety of blood-contact devices

15

because it is involved in prevention of excessive tissue ingrowth (intimal hyperplasia)

16

and thrombogenesis [32, 33]. It was also reported that the negatively charged

17

heparin-mimicking polymer-modified surface was favorable for cell proliferation by 22

1

immobilization of cellular fibronectin and thus exhibited extremely low cytotoxicity

2

[48]. Thus, in this paper, the heparin-mimicking membranes were expected to enhance

3

the cell adhesion ability and be favorable for cell proliferation. To evaluate the cell

4

viability of the membranes, HUVECs were used as model cells.

5

3.3.1. Cell viability and morphology

6

To explore the viability and distribution of the HUVECs on the membrane surfaces,

7

fluorescence images for the live/dead-stained HUVECs on the pristine and

8

heparin-mimicking PES membranes after 6 days incubation were obtained, as shown

9

in Fig.7. The live/dead staining results (Fig.7, A-G) revealed that on the

10

heparin-mimicking membrane surfaces large amounts of live cells were observed (in

11

green) and negligible dead cells were observed (in red). However, no dense confluent

12

cell layers were observed on the pristine PES membrane.

13

Confocal microscopy techniques were carried out to further evaluate the

14

morphologies of the cells seeded onto the membrane surfaces. As shown in Fig.7 (a-g),

15

the amounts of the HUVECs seeded onto the heparin-mimicking membranes were

16

higher compared to pristine PES membrane; and the cells on all the modified

17

membranes exhibited a spread morphology with well-defined actin fibers (in red)

18

throughout the aggregated clusters; however, fewer cells and stress fibers were

19

observed on pristine PES membrane. The HUVEC is a kind of anchorage-dependent

20

cell, which could produce abundant of proteins and spread with ruffling of peripheral

21

cytoplasm. Furthermore, the heparin-mimicking structure was able to bind and

22

stabilize cell growth factor during cell attachment and growth [49], which might also 23

1

promote the cell adhesion and growth on the heparin-mimicking surfaces.

2

The results confirmed that the modified membrane surfaces could promote the

3

HUVECs adhesion and growth, which made them excellent candidates used in blood

4

purification field.

5 6

Fig.7 Fluorescence images (A-G) of live/dead-stained HUVECs and LSCF images

7

(a-g) of HUVECs cultured on the surfaces of the membranes after 6 days

8

3.3.2. MTT assay

9

MTT assay was also carried out to evaluate the cytotoxicity of the membranes. As

10

shown in Fig.8, the absorbance of the formazan increased with increasing the culture

11

time. The absorption of all the modified membranes was higher compared to pristine

12

PES membrane; and with the time prolonged, this trend became more obvious, which

13

indicated a higher viability of the cells for the modified membranes. It was found that

14

the M-2-3 and M-3-2 membranes exhibited the best cytocompatibility among the

15

heparin-mimicking membranes. As reported that the cell−material surface interaction 24

1

was influenced by various factors such as surface charge, surface wettability, free

2

energy, surface morphology, roughness, and the existence of bioactive factors, etc.

3

[48]. Thus, the difference in the MTT data for these samples might be resulted from

4

these complicated factors. In a word, the heparin-mimicking membranes showed a

5

lower cytotoxicity and better cell viability compared to pristine PES membranes.

6 7

Fig.8 MTT assay results. Formazan absorbance was expressed as a function of time

8

for the HUVECs seeded onto different membranes and the control; values are

9

expressed as mean ± SD of 12 determinations, P*, P** and P***< 0.05

10

3.4. Protein antifouling property

11

When membranes were used for hemodialysis, protein molecules would deposit

12

and/or adsorb on the surfaces and the pore surfaces of the membranes, and resulting in

13

membrane fouling. Thus in this work, ultrafiltration of BSA solution was carried out

14

to investigate the antifouling property of the membranes. The results were presented

15

in Fig.9 and Table 2.

16

3.4.1. PBS and BSA solution fluxes for the membranes

17

As shown in Fig.9, all the modified membranes had higher initial PBS fluxes than 25

1

the pristine PES membrane. For the M-2-0, M-2-1, M-2-3 membranes, of which

2

contained the same content of poly (AANa), the fluxes of both PBS and BSA solution

3

increased dramatically with increasing the SSNa component, especially for the M-2-3

4

membrane, and the PBS flux reached 556.04 g/m2·h·mmHg. The flux data were

5

presented in Table 2.

6

In this study, the modified membrane surfaces had higher affinity for water, and

7

consequently lead to the increase in water flux, similar to our previous studies [22, 24].

8

This mainly resulted from the thinner skin layer and macrovoid sub-layer formed

9

during membrane preparation, which could be distinctly observed in Fig.2 (For the

10

M-2-0, M-2-1 and M-2-3 samples, the pore sizes became larger and the porosity

11

increased with increasing the hydrophilic contents). The higher fluxes of the modified

12

membranes had no effect on practical application, since the fluxes of membranes

13

could be effectively adjusted by changing the PES concentration in casting solution or

14

the solvent concentration in bore liquid [50].

15

When the feed solution changed from PBS to BSA solution, the fluxes dramatically

16

decreased (as shown in Fig.9) due to the deposition and adsorption of protein

17

molecules onto the membrane surfaces and/or in the membrane pore surfaces, which

18

had also been observed in many other works [35, 51]. After BSA solution

19

ultrafiltration, the membrane was rinsed in PBS, then the PBS flux was measured

20

again, and the flux recovered to some extent.

26

1 2

Fig.9 Time-dependent fluxes of the membranes at room temperature (PBS: 0–20 min;

3

110–130 min and 220-240 min; BSA solution: 30–50 min and 140-160 min). The

4

membranes were rinsed in PBS for 60 min after BSA solution permeation (values are

5

expressed as mean ± SD of 5 determinations)

6

3.4.2. Recycling property of the membranes

7

The flux recovery ratios could obviously present the suitable recycling properties of

8

the membranes, as shown in Table 2. Generally, a higher JRR value means a better

9

resistance to protein contamination.

10

All the JRR values for the M-2-0, M-2-1 and M-2-3 membranes were more than 90 %

11

and higher than that for the M-0-0 of 50.85 %. This might be resulted from the

12

enhancement of the membrane hydrophilicity. It was reported that a combined water

13

layer would form at the hydrophilic surface, which could further inhibit the protein

14

from adsorbing onto the surface [38, 52]. When the PBS permeated the 27

1

heparin-mimicking membranes for the second time, all the JRR values could even

2

reach over 85 %, which were still higher than that for the M-0-0 membrane; especially

3

for the M-2-0 membrane, the second JRR value could reach 106%. Thus, it could be

4

concluded that the heparin-mimicking membranes showed a better protein antifouling

5

property than the pristine PES membrane.

6

As listed in Table.2, the BSA rejection ratios (R, here the R values were calculated

7

from the first BSA ultrafiltration experiment) for the M-2-0 and M-2-1 membranes

8

were higher than that for membrane M-0-0. However, the rejection ratio reduced

9

when the hydrophilic component increased, especially for the membranes M-2-3 and

10

M-3-2. Table.2. BSA ultrafiltration data for the membranes

11

12

Sample

Initial PBS flux (Jv) (mL/m2·h·mmHg)

Rejection ratio (R)

M-0-0 M-2-0 M-2-1 M-2-3 M-0-2 M-1-2 M-3-2

35.73 66.13 298.26 556.04 134.29 384.64 850.25

88.99 92.34 83.68 71.75 95.84 96.43 56.36

JRR of BSA ultrafiltration 1st time 2nd time 50.85 66.04 90.82 106.67 92.28 88.73 96.76 86.43 99.25 97.00 78.91 92.41 76.82 96.48

3.5. Antibacterial property

13

The antibacterial activity of the modified membranes was firstly tested via bacterial

14

inhibition zone toward E. coli and S. aureus, respectively [53]. As shown in Fig. 8 (a),

15

the pristine PES membrane showed no bacterial inhibition ability, while the Ag

16

nanoparticle loaded membranes showed significant inhibition effect on E. coli. and S.

17

aureus. The sizes of the inhibition zones for E. coli were 0.0 mm (M-0-0), 3.3 mm 28

1

(M-6-0), 2.1 mm (M-4-0), and 1.0 mm (M-2-0), respectively. As for S. aureus, the

2

inhibition zones were 0.0 mm (M-0-0), 4.2 mm (M-6-0), 3.0 mm (M-4-0) and 1.5 mm

3

(M-2-0), respectively. The results indicated that the Ag nanoparticle loaded

4

membranes had significant inhibition capacity toward both gram-negative and

5

gram-positive bacteria.

6

To confirm the antibacterial property in aqueous solution, the optical degree of the

7

bacterial-membrane co-cultured solution was detected. As shown in Fig.10 (b),

8

significant bacterial growth was observed from the control sample and the pristine

9

PES membrane after 4h, 8h and 12 h, respectively. However, the optical degree for the

10

Ag NPs embedded membranes exhibited considerable reduction for both S. aureus

11

and E. coli. The results clearly demonstrated that the Ag NPs embedded membranes

12

had good effect on the inhibition of the growth for both S. aureus (gram positive) and

13

E. coli (gram negative). As revealed by earlier literatures [54, 55] the antibacterial

14

activity of the modified membranes was mainly caused by the released Ag+ ions from

15

the membranes.

29

1 2

Fig.10 (a) The inhibition zone pictures for E. coli (gram negative) and S. aureus (gram

3

positive). (b) The optical degrees for E. coli (gram negative) and S. aureus (gram

4

positive), the absorbance represented the bacterial amount after exposure to the

5

membranes for 4 h, 8 h and 12 h, respectively. The asterisks (*) indicate that the

6

difference attained a statistically significant decrease compared with the control. *P <

7

0.05.

8

3.6. Systematical comparison of different studies

9

To confirm the high efficiency and convenience of the in-situ crosslinking

10

polymerization method, a systematical comparison of some previous studies was

11

made and shown in Table.3. It could conclude that the performance of the membranes

12

prepared from the in-situ cross-linked polymer solution was equal to those prepared

13

from blending method (as ref.14, 16 and 56). In addition, the heparin-mimicking

14

membranes prepared in this work showed longer clotting times, better anti-fouling

15

property, regardless of the higher WCAs and PBS fluxes. Thus, it could conclude that 30

1

the

heparin-mimicking

membranes

coupled

with

the

in-situ

crosslinking

2

polymerization method had great potential to be used in blood purification field. Table.3 Systematical comparison of the results for several studies

3

Membrane Property

Functional macromolecu lar (FM) Fluxes (mL/m2·.h·m mHg) WCAs decrease Blood clotting Anti-protein adsorption Anti-platelet adhesion

This work

[14]

Cross-linke d PAANa and PSSNa

Poly(St-coAA)-blockPoly(VP)-bl ockPoly(St-coAA)

~300-800

No information

++

Anti-Platelet activation 4

* +: positive effect

5

4. Conclusions

[56]

[35]

[16]

MPEG-P(SSNaco-MMA)

Cross-linke d poly(HEM A-co-AA)

Heparin-m imicking PES

No information

~160

No informatio n

+++

++

+++

++

+++

+++

++

++

++

+++

No information

++

++

++

+++

++

++

++

++

++

No information

No information

No information

No informatio n

6

In this study, heparin-mimicking PES membranes were prepared via a high efficient,

7

convenient and universal in-situ cross-linking polymerization coupled with phase

8

inversion technique. The heparin-mimicking membranes showed decreased water

9

contact angle, increased protein antifouling property, prolonged clotting times, and

10

suppressed platelet adhesion compared to pristine membrane. The heparin-mimicking

11

surface also showed better performance in endothelial cells proliferation and cell 31

1

morphology. These results proved that the in-situ heparin-mimicking protocol had

2

endowed the membranes with excellent biocompatibility and antifouling property; and

3

the membranes had great potential to be applied in blood purification fields. In

4

addition, after loading Ag nanoparticles, the membranes exhibited significant

5

inhibition capability for S. aureus and E. coli. All these results proved that the in-situ

6

cross-linked polymerization coupled with phase inversion method would be a

7

versatile protocol for functional modification of polymeric membranes and had great

8

potential in the preparation of membranes used for future portable hemodialyzer.

9 10 11 12 13 14 15 16

Acknowledgments

17

This work was financially sponsored by the National Natural Science Foundation of

18

China (Nos. 51173119, 51225303 and 51433007). We would also thank our laboratory

19

members for their generous help, and gratefully acknowledge the help from Ms. Hui

20

Wang, of the Analytical and Testing Center at Sichuan University, for SEM

21

observation, Ms. Liang of the Department of Nephrology at West China Hospital for

22

the human fresh blood collection. 32

1

References and Notes

2

[1] I. Capila, R.J. Linhardt, Heparin–protein interactions, Angew. Chem., Int. Ed., 41

3

(2002) 390-412.

4

[2] Y. Tamada, M. Murata, T. Hayashi, K. Goto, Anticoagulant mechanism of

5

sulfonated polyisoprenes, Biomaterials, 23 (2002) 1375-1382.

6

[3] I.A. Perrenoud, E.C. Rangel, R.P. Mota, S.F. Durrant, N.C.d. Cruz, Evaluation of

7

blood compatibility of plasma deposited heparin-like films and SF6 plasma treated

8

surfaces, Mater. Res., 13 (2010) 95-98.

9

[4] K. Ichinose, T. Okamoto, H. Tanimoto, A. Yoshitake, M. Tashiro, Y. Sakanashi, K.

10

Kuwana, K. Tahara, M. Kamiya, H. Terasaki, Comparison of a New Heparin‐coated

11

Dense Membrane Lung with Nonheparin ‐ coated Dense Membrane Lung for

12

Prolonged Extracorporeal Lung Assist in Goats, Artif. Organs, 28 (2004) 993-1001.

13

[5] R. Kopp, K. Mottaghy, M. Kirschfink, Mechanism of complement activation

14

during extracorporeal blood-biomaterial interaction: effects of heparin coated and

15

uncoated surfaces, ASAIO J., 48 (2002) 598-605.

16

[6] W.-C. Lin, T.-Y. Liu, M.-C. Yang, Hemocompatibility of polyacrylonitrile dialysis

17

membrane immobilized with chitosan and heparin conjugate, in:

18

pp. 1947-1957.

19

[7] D. Labarre, Heparin-like polymer surfaces: control of coagulation and

20

complement activation by insoluble functionalized polymers, The International

21

journal of artificial organs, 13 (1990) 651-657.

22

[8] R. Mammadov, B. Mammadov, S. Toksoz, B. Aydin, R. Yagci, A.B. Tekinay, M.O.

23

Guler,

24

Biomacromolecules, 12 (2011) 3508-3519.

25

[9] Ø. Arlov, F.L. Aachmann, A. Sundan, T. Espevik, G. Skjåk-Bræk, Heparin-Like

26

Properties of Sulfated Alginates with Defined Sequences and Sulfation Degrees,

27

Biomacromolecules, 15 (2014) 2744-2750.

28

[10] J. Dernedde, A. Rausch, M. Weinhart, S. Enders, R. Tauber, K. Licha, M.

29

Schirner, U. Zügel, A. von Bonin, R. Haag, Dendritic polyglycerol sulfates as

30

multivalent inhibitors of inflammation, Proc.Natl.Acad.Sci., 107 (2010) 19679-19684.

Heparin

mimetic

peptide

nanofibers

33

Biomaterials, 2004,

promote

angiogenesis,

1

[11] K.L. Christman, V. Vázquez-Dorbatt, E. Schopf, C.M. Kolodziej, R.C. Li, R.M.

2

Broyer, Y. Chen, H.D. Maynard, Nanoscale growth factor patterns by immobilization

3

on a heparin-mimicking polymer, J. Am. Chem. Soc., 130 (2008) 16585-16591.

4

[12] L. Li, C. Cheng, T. Xiang, M. Tang, W. Zhao, S. Sun, C. Zhao, Modification of

5

polyethersulfone hemodialysis membrane by blending citric acid grafted polyurethane

6

and its anticoagulant activity, J. Membr. Sci., 405 (2012) 261-274.

7

[13] S. Nie, J. Xue, Y. Lu, Y. Liu, D. Wang, S. Sun, F. Ran, C. Zhao, Improved blood

8

compatibility of polyethersulfone membrane with a hydrophilic and anionic surface,

9

Colloid Surf. B-Biointerfaces, 100 (2012) 116-125.

10

[14] F. Ran, S. Nie, J. Li, B. Su, S. Sun, C. Zhao, Heparin‐Like Macromolecules for

11

the Modification of Anticoagulant Biomaterials, Macromol. Biosci., 12 (2012)

12

116-125.

13

[15] M. Tang, J. Xue, K. Yan, T. Xiang, S. Sun, C. Zhao, Heparin-like surface

14

modification of polyethersulfone membrane and its biocompatibility, J. Colloid

15

Interface Sci., 386 (2012) 428-440.

16

[16] S. Nie, M. Tang, Z. Yin, L. Wang, S. Sun, C. Zhao, Biologically inspired

17

membrane design with a heparin-like interface: prolonged blood coagulation,

18

inhibited complement activation, and bio-artificial liver related cell proliferation,

19

Biomater. Sci., 2 (2014) 98-109.

20

[17] L. Wang, H. Qin, S. Nie, S. Sun, F. Ran, C. Zhao, Direct synthesis of heparin-like

21

poly (ether sulfone) polymer and its blood compatibility, Acta Biomater., 9 (2013)

22

8851-8863.

23

[18] W. Zhao, Q. Mou, X. Zhang, J. Shi, S. Sun, C. Zhao, Preparation and

24

characterization of sulfonated polyethersulfone membranes by a facile approach, Eur.

25

Polym. J., 49 (2013) 738-751.

26

[19] F. Ran, S. Nie, Z. Yin, J. Li, B. Su, S. Sun, C. Zhao, Synthesized negatively

27

charged macromolecules (NCMs) for the surface modification of anticoagulant

28

membrane biomaterials, Int. J. Biol. Macromol., 55 (2013) 269-275.

29

[20] B. Nandan, K. Pandey, G. Pandey, A. Singh, L. Kandpal, G. Mathur, Miscibility

30

behaviour of poly (ether sulphone)/nylon-6 blends, J. Therm. Anal. Calorim., 64 34

1

(2001) 529-537.

2

[21] V. Deimede, D. Labou, S.G. Neophytides, Polymer electrolyte membranes based

3

on blends of sulfonated polysulfone and PEO‐grafted polyethersulfone for low

4

temperature water electrolysis, J. Appl. Polym. Sci., 131 (2014).

5

[22] Z. Han, C. Cheng, L. Zhang, C. Luo, C. Nie, J. Deng, T. Xiang, C. Zhao, Toward

6

robust pH-responsive and anti-fouling composite membranes via one-pot in-situ

7

cross-linked copolymerization, Desalination, 349 (2014) 80-93.

8

[23] Z.-Q. Shi, X.-L. Huang, C. Wang, Y.-F. Li, C. He, C.-S. Zhao, In Situ

9

Cross-Linked Polymerization toward Poly (ether sulfone)/Poly (sodium acrylate)

10

Hybrid Particles for the Removal of Environmental Toxins, Ind. Eng. Chem. Res., 53

11

(2014) 14084-14093.

12

[24] T. Xiang, L.-R. Wang, L. Ma, Z.-Y. Han, R. Wang, C. Cheng, Y. Xia, H. Qin,

13

C.-S. Zhao, From commodity polymers to functional polymers, Sci. Rep., 4 (2014).

14

[25] J. Lu, Z. Qiu, W. Yang, Fully biodegradable blends of poly (l-lactide) and poly

15

(ethylene succinate): miscibility, crystallization, and mechanical properties, Polymer,

16

48 (2007) 4196-4204.

17

[26] C. Cheng, S. Sun, C. Zhao, Progress in heparin and heparin-like/mimicking

18

polymer-functionalized biomedical membranes, J. Mater. Chem. B, 2 (2014)

19

7649-7672.

20

[27] Z. Chang‐sheng, L. Ting, L. Zhong‐ping, C. Li‐ping, H. Jia, An evaluation

21

of a polyethersulfone hollow fiber plasma separator by animal experiment, Artif.

22

Organs, 25 (2001) 60-63.

23

[28] X.J. Huang, D. Guduru, Z.K. Xu, J. Vienken, T. Groth, Blood Compatibility and

24

Permeability of Heparin ‐ Modified Polysulfone as Potential Membrane for

25

Simultaneous Hemodialysis and LDL Removal, Macromol. Biosci., 11 (2011)

26

131-140.

27

[29] Y. Li, C. Cao, T.-S. Chung, K.P. Pramoda, Fabrication of dual-layer

28

polyethersulfone (PES) hollow fiber membranes with an ultrathin dense-selective

29

layer for gas separation, J. Membr. Sci., 245 (2004) 53-60.

30

[30] Q. Yang, T.-S. Chung, S.B. Chen, M. Weber, Pioneering explorations of rooting 35

1

causes for morphology and performance differences in hollow fiber kidney dialysis

2

membranes spun from linear and hyperbranched polyethersulfone, J. Membr. Sci., 313

3

(2008) 190-198.

4

[31] C. Zhao, J. Xue, F. Ran, S. Sun, Modification of polyethersulfone membranes–a

5

review of methods, Prog. Mater Sci., 58 (2013) 76-150.

6

[32] A. de Mel, G. Jell, M.M. Stevens, A.M. Seifalian, Biofunctionalization of

7

biomaterials for accelerated in situ endothelialization: a review, Biomacromolecules, 9

8

(2008) 2969-2979.

9

[33] Y.M. Shin, Y.B. Lee, S.J. Kim, J.K. Kang, J.-C. Park, W. Jang, H. Shin,

10

Mussel-inspired immobilization of vascular endothelial growth factor (VEGF) for

11

enhanced endothelialization of vascular grafts, Biomacromolecules, 13 (2012)

12

2020-2028.

13

[34] V. Hlady, J. Buijs, Protein adsorption on solid surfaces, Curr. Opin. Biotechnol., 7

14

(1996) 72-77.

15

[35] H. Qin, C. Sun, C. He, D. Wang, C. Cheng, S. Nie, S. Sun, C. Zhao, High

16

efficient protocol for the modification of polyethersulfone membranes with

17

anticoagulant and antifouling properties via in situ cross-linked copolymerization, J.

18

Membr. Sci., 468 (2014) 172-183.

19

[36] C. Cheng, S. Nie, S. Li, H. Peng, H. Yang, L. Ma, S. Sun, C. Zhao, Biopolymer

20

functionalized reduced graphene oxide with enhanced biocompatibility via mussel

21

inspired coatings/anchors, J. Mater. Chem. B, 1 (2013) 265-275.

22

[37] T. Xiang, W.-W. Yue, R. Wang, S. Liang, S.-D. Sun, C.-S. Zhao, Surface

23

hydrophilic modification of polyethersulfone membranes by surface-initiated ATRP

24

with enhanced blood compatibility, Colloid Surf. B-Biointerfaces, 110 (2013) 15-21.

25

[38] Q. Wei, T. Becherer, S. Angioletti‐Uberti, J. Dzubiella, C. Wischke, A.T. Neffe,

26

A. Lendlein, M. Ballauff, R. Haag, Protein interactions with polymer coatings and

27

biomaterials, Angew. Chem., Int. Ed., 53 (2014) 8004-8031.

28

[39] Y. Xia, C. Cheng, R. Wang, H. Qin, Y. Zhang, L. Ma, H. Tan, Z. Gu, C. Zhao,

29

Surface-engineered nanogel assemblies with integrated blood compatibility, cell

30

proliferation and antibacterial property: towards multifunctional biomedical 36

1

membranes, Polym. Chem., 5 (2014) 5906-5919.

2

[40] P.-C. Chen, L.-S. Wan, Z.-K. Xu, Bio-inspired CaCO 3 coating for

3

superhydrophilic hybrid membranes with high water permeability, J. Mater. Chem., 22

4

(2012) 22727-22733.

5

[41] W. Zhao, J. Huang, B. Fang, S. Nie, N. Yi, B. Su, H. Li, C. Zhao, Modification of

6

polyethersulfone membrane by blending semi-interpenetrating network polymeric

7

nanoparticles, J. Membr. Sci., 369 (2011) 258-266.

8

[42] B. Fang, C. Cheng, L. Li, J. Cheng, W. Zhao, C. Zhao, Surface modification of

9

polyethersulfone membrane by grafting bovine serum albumin, Fiber. Polym., 11

10

(2010) 960-966.

11

[43] A. Nabe, E. Staude, G. Belfort, Surface modification of polysulfone ultrafiltration

12

membranes and fouling by BSA solutions, J. Membr. Sci., 133 (1997) 57-72.

13

[44] D. Öner, T.J. McCarthy, Ultrahydrophobic surfaces. Effects of topography length

14

scales on wettability, Langmuir, 16 (2000) 7777-7782.

15

[45] J. Grunkemeier, W. Tsai, T. Horbett, Co-adsorbed fibrinogen and von Willebrand

16

factor augment platelet procoagulant activity and spreading, J. Biomater. Sci. Polym.

17

Ed., 12 (2001) 1-20.

18

[46] J. Li, X.J. Huang, J. Ji, P. Lan, J. Vienken, T. Groth, Z.K. Xu, Covalent Heparin

19

Modification of a Polysulfone Flat Sheet Membrane for Selective Removal of Low‐

20

Density Lipoproteins: A Simple and Versatile Method, Macromol. Biosci., 11 (2011)

21

1218-1226.

22

[47] C. Cheng, S. Li, S. Nie, W. Zhao, H. Yang, S. Sun, C. Zhao, General and

23

biomimetic approach to biopolymer-functionalized graphene oxide nanosheet through

24

adhesive dopamine, Biomacromolecules, 13 (2012) 4236-4246.

25

[48] J. Deng, X. Liu, L. Ma, C. Cheng, W. Shi, C. Nie, C. Zhao, Heparin-Mimicking

26

Multilayer Coating on Polymeric Membrane via LbL Assembly of Cyclodextrin based

27

Supramolecules, ACS Appl. Mater. Inter., (2014).

28

[49] T.H. Nguyen, S.-H. Kim, C.G. Decker, D.Y. Wong, J.A. Loo, H.D. Maynard, A

29

heparin-mimicking polymer conjugate stabilizes basic fibroblast growth factor, Nat.

30

Chem., 5 (2013) 221-227. 37

1

[50] N. Peng, N. Widjojo, P. Sukitpaneenit, M.M. Teoh, G.G. Lipscomb, T.-S. Chung,

2

J.-Y. Lai, Evolution of polymeric hollow fibers as sustainable technologies: past,

3

present, and future, Prog. Polym. Sci., 37 (2012) 1401-1424.

4

[51] L. Li, Z. Yin, F. Li, T. Xiang, Y. Chen, C. Zhao, Preparation and characterization

5

of poly (acrylonitrile-acrylic acid-N-vinyl pyrrolidinone) terpolymer blended

6

polyethersulfone membranes, J. Membr. Sci., 349 (2010) 56-64.

7

[52] H. Chen, L. Yuan, W. Song, Z. Wu, D. Li, Biocompatible polymer materials: role

8

of protein–surface interactions, Prog. Polym. Sci., 33 (2008) 1059-1087.

9

[53] X. Cao, C. Cheng, Y. Ma, C. Zhao, Preparation of silver nanoparticles with

10

antimicrobial activities and the researches of their biocompatibilities, J. Mater. Sci.:

11

Mater. Med., 21 (2010) 2861-2868.

12

[54] I. Sondi, B. Salopek-Sondi, Silver nanoparticles as antimicrobial agent: a case

13

study on E. coli as a model for Gram-negative bacteria, J. Colloid Interface Sci., 275

14

(2004) 177-182.

15

[55] M. Rai, A. Yadav, A. Gade, Silver nanoparticles as a new generation of

16

antimicrobials, Biotechnol. Adv., 27 (2009) 76-83.

17

[56] C. He, C.X. Nie, W.F. Zhao, L. Ma, T. Xiang, C.S. Cheng, S.D. Sun, C.S. Zhao,

18

Modification of polyethersulfone membranes using terpolymers engineered and

19

integrated antifouling and anticoagulant properties, Polym. Adv. Technol., 24 (2013)

20

1040-1050.

21 22

38

1

Captions for scheme and figures

2 3

Scheme.1 Preparation process and multi-functionality of the heparin-mimicking

4

membranes.

5 6

Fig.1 ATR-FTIR spectra for the membranes

7 8

Fig.2 SEM micrographs for (a) the cross-section views and (b) the surface views of

9

the membranes. (Magnification: ×1000; Scale bar: 10μm)

10 11

Fig.3 Static water contact angle for the pristine and heparin-mimicking PES

12

membranes

13 14

Fig.4 BSA and FBG adsorption for the pristine and heparin-mimicking PES

15

membranes

16 17

Fig.5 (a) SEM images of the platelets adhering onto the membranes, images (a-g)

18

were the partial enlarged ones for images (A-G); (b) the number of the adhering

19

platelets onto the membranes (from platelet-rich plasma estimated by SEM images)

20 21

Fig.6 APTTs and TTs for the pristine PES and the heparin-mimicking membranes

22

(Values are expressed as mean ± SD. P#, P* < 0.05 compared with plasma (PPP) and 39

1

pristine PES membrane, respectively.)

2 3

Fig.7 Fluorescence images (A-G) of live/dead-stained HUVECs and LSCF images

4

(a-g) of HUVECs cultured on the surfaces of the membranes after 6 days

5 6

Fig.8 MTT assay results. Formazan absorbance was expressed as a function of time

7

for the HUVECs seeded onto different membranes and the control; values are

8

expressed as mean ± SD of 12 determinations, P*, P** and P***< 0.05

9 10

Fig.9 Time-dependent fluxes of the membranes at room temperature (PBS: 0–20 min;

11

110–130 min and 220-240 min; BSA solution: 30–50 min and 140-160 min). The

12

membranes were rinsed in PBS for 60 min after BSA solution permeation (values are

13

expressed as mean ± SD of 3 determinations)

14 15

Fig.10 (a) The inhibition zone pictures for E. coli (gram negative) and S. aureus (gram

16

positive). (b) The optical degrees for E. coli (gram negative) and S. aureus (gram

17

positive), the absorbance represented the bacterial amount after exposure to the

18

membranes for 4 h, 8 h and 12 h, respectively. The asterisks (*) indicate that the

19

difference attained a statistically significant decrease compared with the control. *P <

20

0.05.

21 22

Highlights: 40

1 2



Heparin-mimicking membranes are prepared using sodium acrylate and sodium styrene sulfonate.

3 4 5



The membranes show excellent blood compatibility and cytocompatibility.



The resulted membranes show enhanced protein antifouling property.



Further modification using Ag nanoparticles endow the samples with good

6 7 8 9 10

antibacterial property.

11

41